Although insertional mutagenesis in transgenic mice has allowed a number of new mutants to be identified and studied, the overall insertion frequency per locus is quite low. Our findings with two transgenic strains, 412 and 413, carrying multiple transgenes (proviruses) confirm previous estimates that only 10% of integrations of transgenes lead to recessive developmental mutations. It has been estimated that the probability of generating a mutation by transgene insertion is 100 fold lower than the probability of producing a mutation using ethylnitrosourea, a potent chemical mutagen. This estimate was based on transgenic strains usually having only one site of integration of exogenous DNA. While screening animals with only one or a few insertions seems theoretically desireable, because large numbers of insertions might make it difficult to determine which one causes the mutation, the low mutagenesis frequency makes large scale insertional mutagenesis with conventional transgenic strains impractical. Although the frequency of insertional mutagenesis cannot be changed, the efficiency can be increased by screening transgenics with multiple insertions, as we have shown with the 412 and 413 strains. We have developed an approach to improve the efficiency and applicability of this method even more, and thereby make it possible to identify and clone a large number of developmentally important genes. We are now implementing this strategy.